Prediction #1
The era of proving the value of your AI investment is now

AI budgets are booming, but the demand for tangible value is catching up. In 2026, the organizations that thrive will treat AI spending like any strategic investment: measurable, governed and accountable.

As AI oversight frameworks expand across North America, EU and APAC, companies will face mounting pressure to demonstrate value alignment between their AI initiatives and business outcomes. Overspending on unproven models will no longer fly.

CIOs and CDAOs will be tasked with balancing innovation against risk, ensuring that each AI project meets compliance, transparency and ROI criteria before launch. Prudent AI investment is not about spending less. It’s about spending smarter.

By 2028, as organizations struggle to understand the actual costs and value of implementations, 65% of "build your own" agentic AI projects will be discontinued after falling short of ROI goals.

Source: "IDC FutureScape: Worldwide Agentic Artificial Intelligence 2026 Predictions, US53860925, October 2025"’

Collibra POV:

Instead of slowing innovation, AI regulation ensures innovation persists responsibly. Collibra AI Governance helps organizations evaluate, document and monitor every AI use case, so data leaders can show measurable return while staying compliant.

In 2025, many leaders invested in AI and were surprised that money did not materialize. They fell into the common trap of leading with tech investment instead of business case. As organizations mature in AI and more consistently view it as a technology that helps them build data assets, derive insights, and automate action for very specific, lucrative goals — and manage the value realization from achieving those goals—the returns will come.

Dr. Barbara Wixom, Principal Research Scientist, MIT CISR

How confident are you that your organization’s AI spend is delivering measurable ROI?

Very confident — we track ROI on all AI projects
Somewhat confident — we have partial visibility
Unsure — ROI is anecdotal at best
Not confident — governance gaps make measurement difficult